NAR Housing Affordability Index is Worthless
The monthly update to the NAR Housing Affordability Index gets released Thursday morning (August 14), as well as the Quarterly Housing Affordability Index for First-time Buyers.
Some people seem to think this index is meaningful. Over the past month, I have received numerous emails explaining to me how "affordable" Housing has become, most notably via this index.
To determine as to whether that was true or not, we looked more closely into the Housing Affordability Index (link,
or AFFDCMOM on your Bloomberg terminal) -- its methodology, what it contains and in particular, what it omits.
Our conclusion? The index as presently constructed is utterly worthless. It provides little or no insight into how affordable US Housing actually is. Further, what is omitted from the index is especially relevant to the problems occurring in the housing market today. The Index fails to account for -- or even recognize -- any of the out of the ordinary circumstances that are currently bedeviling the Hosuing market.
Consider the red line in the nearby chart (click for larger graphic). That is the NAR HAI rating of "100" -- what the NAR states as their baseline measure of affordability. As hard as this might be to imagine, it shows that over the course of the biggest run up in housing prices in American history, the Index remained perfectly affordable. Except for one monthly reading of 99.55 in late 2005 -- a smidge below 100 -- housing never dipped into the level of unaffordable over the entire giant housing boom.
This is mind bogglingly astonishing. If the affordability index failed to show housing was unaffordable during 2005-06, when would it ever show that?
Given this rather extremely dubious conclusion, we simply had to look at how the Index was composed, to see if we can figure out where it went so terribly astray. We conclude that the index is overly simple, that it fails to include many key factors of the current financial crisis. The index ignores factors like family savings rates, available cash assets, consumer credit, indebtedness, credit servicing obligations inflation, income gains, and mortgage availability.
These are crucial factors impacting the current housing situation. Hence, why today's missive will caution you against putting any weight whatsoever on the NAR Housing Affordability Index. It is, to be blunt, without any value at all.
Let's take a closer look at what goes into it, and what's missing, so you might better understand its weaknesses and flaws.
The source: First, I begin with the assumption that readers are aware of the National Association of Realtors abysmal track record. Throughout the entire downturn, they have acted not as objective purveyors of data, but as cheerleaders and spin doctors. (See Tracking NAR Spin, or Worst. Forecasters. Ever? or NAR and Housing Forecasts for the ugly details). The bottom line is that one should treat any analysis from the NAR warily, and look very hard at their data and forecasts.
Note: This is not a mere ad hominem attack, but rather, it is simply a case of accountability for the reports and commentaries of the past 3 years. They have been wildly wrong for an extensive period of time.
Onto the Index: The NAR Housing Affordability Index is quite simple in its construction: The real estate broker’s association looks at whether the median family income “qualifies” for the median house price using prevailing interest rates. Qualifying ratio of 25% of the median income has to cover 100% of the monthly mortgage P&I, which is 80% of the home's purchase price.
Take the median home price as calculated by the NAR. Use the prevailing interest rates, as determined by via Federal Housing Finance Board. Then calculate how much principle and interest of a monthly mortgage for 80% of that property value would be. Lastly, take the median national income (via U.S. Bureau of the Census).
A value of 100 means that a family with the median income has exactly enough income to qualify for a mortgage on a median-priced home. 120 = 120% of the income necessary, 80 = 80% of the income necessary. (See: Methodology for the Housing Affordability Index).
Voila! Affordability Index!
Where things get interesting is when we look into some of the assumptions of the index, and what was omitted from consideration.
Consider the omissions:
1) The NAR methodology "assumes" a cash down payment of 20%.
Given the negative savings rate and enormous extension of credit we've seen over the past 6 years, far fewer families have that 20% down available than has been the case in the past.
2) The NAR “assumes” 25% of monthly gross income is available to pay for Mortgage Principle and Interest.
The Index ignores the current amount of debt being carried by the typical household.
Given the increase of debt (both HELOC and Revolving) that has occurred, the debt service is also much greater. Hence, that’s a big assumption, and one that explicitly ignores the current levels of household debtload. The consumer is extended, and their income has not kept up with inflation over the past 6 years. Hence, they may not have 25% of their income available for a mortage if it is already committed to servicing other debts. Banks and brokers aren't the only ones who need to deleverage...
3) The Index fails to consider FICO scores.
Beyond the total amount of debt, the next flaw in the index is the FICO credit score.
As we have seen, late payments are on the rise for everything from credit cards, to existing mortgages, to auto payments, and student loans. Hence, we would expect to see FICO scores moving downwards, especially amongst middle and lower income potential home buyers. This will especialyl impact buyers of "Starter homes."
4) An increasing number of families no longer qualify for a standard conforming mortgage.
All of the above leads us to recognize that prevailing interest rates are less relevant for many buyers . Given their debt levels and FICO scores, many will end up in Alt-A -- or worse. Assuming that option is even available.
5) Non purchase costs of home ownership have risen dramatically .
All of these factors only go to buying a house; they fail to consider the increased costs of owning a home. The past few years have seen a big rise in Property Taxes, a large increase in maintenance costs, and an enormous increase in heating Oil, etc. These elements are not considered relevant to the affordability index.
6) Regional variations in home prices and income are ignored.
It most be noted that the index is a national index. It fails to consider regional variations in home prices, industries, incomes, etc. Thus, the index does not take into consideration that some regions of the nation are far more affordable than others.
Bottom line: There isn’t a whole lot of affordability built into the NAR affordability index.
~~~
Let's do a more thoughtful housing analysis -- one that reviews home affordability factor according to other variables -- the ratio between median income and median home price, or the rental price gains to home appreciation ratio. (See the IMF charts here, and Ned Davis and ISI below). These make sense, as they determine a) if it makes sense for renters to buy; and 2) If renters can afford to buy and/or if existing owners can afford to trade up.
Median New Home Prices vs Median Household Disposable Income
Home Price Appreciation vs CPI - Rent
Charts courtesy of Ned Davis Research via Comstock Funds
Previously:
NAR and Housing Forecasts (June 2007)
http://bigpicture.typepad.com/comments/2007/06/nar_and_housing.html
Tracking NAR Spin (April 2008)
http://bigpicture.typepad.com/comments/2008/04/tracking-nar-sp.html
Sources:
Housing Affordability Index
http://www.realtor.org/research/research/housinginx
Methodology for the Housing Affordability Index
http://www.realtor.org/research/research/hameth
Wednesday, August 13, 2008 | 09:00 AM | Permalink
| Comments (37)
| TrackBack (0)
add to de.li.cious | 
digg this! | 
add to technorati | 
email this post
TrackBack
TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d8341c52a953ef00e553bf3e198833
Listed below are links to weblogs that reference NAR Housing Affordability Index is Worthless:
Comments
Awesome analysis BR with all the factors that most ignore. How about price of houses rocketing up without wage increases? Most people I know are making the same or less than they were 8 years ago.
This presentation is pretty good too just looking at prices during a bubble. Chris feels it should be awhile for the prices to bottom.
http://www.chrismartenson.com/bubbles
Posted by: GB | Aug 13, 2008 9:14:41 AM
The NAR's methodology wouldn't be all that bad if they did the analysis by region and then averaged it for the country.
Median house prices don't buy you an awful lot of housing in, say: NYC, LA, Chicago...
but there's plenty of affordable housing in flyover country... if you happen to have a job that pays big city wages.
Posted by: Patrick (G) | Aug 13, 2008 9:35:31 AM
Assuming 20% down is a joke, especially for first time homebuyers. I don't know anybody who put anything close to 20% down. Where I'm looking to buy, 20% of a starter home would be 50 to 60 grand.
Posted by: Renting in Mass | Aug 13, 2008 9:55:57 AM
I believe the term is "endogeneity." If asset prices rise with lower interest rates, and payments fall with lower interest rates - and both statements are largely true - it doesn't tell you much at all, since one compensates the other. If the relationship was perfect, it would actually be a truism.
I'm a little disappoint in the NAR - it should be possible to construct an affordability index that stays at precisely 100 all the time, and I'm disappointed they didn't refine their methodology further.
Posted by: Greg | Aug 13, 2008 9:59:00 AM
Great analysis. If this makes sense to you, check out this tool found at www.UsHousingMeltdown.org where you can type in your zip-code and get the median income, median home price ratios. Look for the Ceiling fundamental.
Posted by: Ryan | Aug 13, 2008 10:08:06 AM
So, after inflation ex-inflation, we now have affordability ex-affordability?
To quote Yves Smith: "Quelle Surprise!"
Posted by: Francois | Aug 13, 2008 10:26:03 AM
The NAR should hire Ben Stein.
Posted by: Unsympathetic | Aug 13, 2008 10:29:01 AM
My wife was looking at houses in lower Fairfield County, CT last week. A house on the market for $1.2 struck her fancy. She asked the broker what a house of this size. location, etc. should cost to rent.
Answer: $4500/month
So let's do the math. If I put down 20%, my mortgage is $960,000. Web sites claim to offer me a 30-year fixed at 6.5% (fat chance I say, but give 'em the benefit of the doubt). So I'm gonna pony up about $6100/month. Taxes are $8000, so another $700 to pay, plus whatever insurance is gonna run me. Not to mention the cost of money on my $240,000 down payment.
So I can buy a depreciateing asset for about $7000/month, or avoid the headache and rent for $4500. Hmmm, affordable my ass.
Posted by: shoeless | Aug 13, 2008 10:30:05 AM
Assuming 20% down is a joke, especially for first time homebuyers.
That's part of the problem, now isn't it ?
Posted by: Patrick (G) | Aug 13, 2008 10:36:37 AM
quote above - "a family with the median income has exactly enough income to qualify for a mortgage"
with all this computing power available
and unemployeed paper pushers
how about diagnosis with
QUINTILES
Posted by: Greg0658 | Aug 13, 2008 11:00:14 AM
Barry, you're applying a rational critique (did a great job btw) to an index generated and marketed by an associated with a vested interested of getting people to buy homes. They are NEVER going to tell people it's a lousy time to buy a home, NEVER. Ever heard of a used car salesman telling a potential customer that it's a lousy time to buy a used car, I haven't. I strongly suspect they knew it was flawed, but accuracy wasn't the point. Marketing value was. The real benefit of your analysis is that perhaps it will go a bit further along the path enlightening some of the bulltards in the MSM of this and they will stop giving the NAR so much unwarranted credibility and publicity.
Posted by: Stuart | Aug 13, 2008 11:15:46 AM
Excellent analysis: insightful and informative. Great job, Barry. I agree that the HAI should be calculated by region, actually, it would be more accurate and relevant if the index were caluculated by MSA. But, hey, we all know that the NAR isn't concerned with accuracy or reliability of its data- the NAR is solely focused on moving units! Just wait until the NAR actually abandons its policy of talking up the market and then shifts its focus to increasing their cut of the sales price. Imagine- a 10% sales commission! Just wait, it'll come sooner or later . . . btw, I absolutely agree with the above post characterizing the realtors as merely used car salesmen!
Posted by: AJF | Aug 13, 2008 11:55:41 AM
Unsympathetic- The monthly nut on a house should never be less than rent so if you are waiting for that you will likely be renting forever. A more accurate comparison would be an interest only mortgage vs. rent. In your calculations you forgot to account for the tax deduction on mortgage interest and propoerty taxes that will account for up to 30% discount to the monthly nut. You also need to take into account that a traditional mortgage is fixed and is a hedge against inflation while rent will generally increase annually. Because of this dynamic the rent vs. buy decision is largely dependent on the amount of time you expect to live in the same place.
You also need to factor the probability of refinancing at a lower interest rate at some point in the future, the transaction costs with a purchase/sale and the expected appreciation or depreciation of the asset until sale.
Comparing the costs of renting vs. buying without giving credit for equity being paid down each month, tax benefits, and the long run hedge against inflation is really just as deceptive as the NAR spin.
Unfortunately the scenario is not a simple rent vs. buy decision. It is usually a "when is the optimal time to buy" decision which requires knowledge of future house prices, future rents, and future interest rates. It is nearly impossible to predict any of those alone with temporal accuracy so good luck predicting all three.
Posted by: R. Timm | Aug 13, 2008 12:05:47 PM
Come on Barry, why you gotta be such a hater? :)
These people are just trying to make an honest buck. It always just comes down to making the numbers work doesn't it?
Great job!
Posted by: shayre | Aug 13, 2008 12:15:28 PM
"Rocky Mountain [Way] Low....."
(apologies to Joe Walsh)
By John Rebchook (Rocky Mountain News)
Tuesday, August 12, 2008
Almost 40 percent of the homes sold in the Denver area in the past year ended in June were sold for a loss, according to a national report released Tuesday.... (cont)
www.rockymountainnews.com/news/2008/aug/12/
for-sellers-its-a-lose-lose-deal/
________________________________________
A sample of neighborhood activity during the second quarter.
Neighborhood Median Year-over-year price change
Baker $246,270 -7.4%
Cherry Creek $591,433 0.6%
Five Points $271,713 Unchanged
Highland $288,035 3.5%
Jefferson Park $274,595 2.2%
Montbello $157,939 -8.1%
North Aurora $121,364 -16.9%
Park Hill $266,473 -1.6
Stapleton $388,368 -1.5%
Washington Park $434,934 1.2%
Posted by: SR | Aug 13, 2008 12:21:27 PM
Here is where it is helpfull and VERY IMPORTANT. CALIFORNIA. Subprime and Alt A loans ARE GONE, PAY OPTION LOANS ARE GONE. What is left. Conforming and FHA.
Considering property taxes and Insurance the index is not that far off.
Now look at history. The Only time Califoria dipped below 20% was 1986 and then it rose back up to 40%, Then started coming down to recently 12% in CA. Why, prices increased due to stated loans, Option Arms, and Interest only with higher ratio's. THE HIGH RATIO LOANS ARE GONE FOR THE MOST PART.
SO NOW YOU CAN CALCUTLATE HOW MUCH THE VALUES OF THE PROPERTIES HAVE TO CORRECT. MY ASSUMPTION WITH THE REMAINING PROGRAMS IS 30% AFFORTABILITY.
Where it also would have been helpfull in the past but the deliberitly stupid politians and investment bankers ignored it was it showed what was going to happen. Prices got so high that reasonable stated incomes would not have qualified.
LEARN FROM THE PAST OR YOU ARE DOOMED TO REPEAT IT. Guess what, CA is and drapgged the rest of the country with it.
dont throw out the baby with the bath water. It was the simplest predicter of what was going to happen.
Posted by: crash | Aug 13, 2008 12:28:19 PM
Lest we forget, the Real Estate industry is the single largest advertiser in newspapers.
So will the papers continue to quote the NAR as a reputable source? ;-)
Posted by: TomD | Aug 13, 2008 1:15:35 PM
Barry
Unlike everyone above, I would tend to disagree that the statistic provided by the NAR is worthless. I agree that using a national average is disengenious, and would prefer to have a stat by MSA.
The stat is great if you look at it this way: for well qualified purchasers, is housing affordable.
What analysis is needed, and where you fall short is as follows
We need a stat - and it should coincide by the same MSA as the NAR stat is THE POPULATION OF HOUSEHODLS/WAGE EARNERS THAT ACTUALLY MEET THE QUALIFICATIONS OF "WELL QUALIFIED PURCHASERS"
That is the banana. That stat takes into account just how many in the msa area can at a given time step up to the plate. The NAR can have there stat that houses for well qualified purchases are becoming more affordable
Our stat provides the clues as to how many purchasers are actually out there, and thus what future sales/inventory absorbtion rates might actually end up being
Kind Regards
Posted by: Jeffrey Levin | Aug 13, 2008 1:27:50 PM
Geez,
Bearing in mind this is a national index (and therefore will have regional flaws), I frankly think they've hit the three key elements correctly.
Median cost of a home - check.
Median family income - check.
Typical cost of credit - check.
Ok, so they made a couple of assumptions - BIG DEAL. We used to do that in the 1970s and 1980s. Regarding down payments - 20% is a long-standing usual for a conventional first mortgage. Check, nothing unusual there.
Assumption regarding amount available to pay the mortgage - 25%, check - guess what - your banker makes a similar assumption. Check, nothing unusual there.
Other than the valid (in my view) criticism about the inflation around other housing cost components (realty taxes, maintenance, heating/cooling costs, etc.),the rest of your argument is fluff Barry.
Does the fact that the consumer has loaded up on credit-based fat-assed trucks, SUVs and over-sized cars make "housing" less affordable? Does the fact that cell phone costs, the movie channel, plasma TVs, home renovations, vacations around globe - all financed on credit - make "housing" less affordable?
No. Only the rise in the cost of non-discretionary items (i.e. food, basic transportation) and those things directly associated with housing might be validy stated to affect the "affordability" of housing.
The "more thoughtful" analysis actually has little to do with absolute homeownership affordability - one deals with the transient relative value of renting versus owning (without considering long-term capital gains either), and the other to do with new home prices versus median income - new housing comprises a very small portion of the total residential realty market.
I'm no fan of the NAR's constant cheerleading of realty, but I don't think most of your criticisms hold much water.
I suppose I'll get flamed by all the Barry cheerleaders now ....
~~~
BR: Jay, how do you reconcile the index not showing homes as unaffordable the entire boom period except one month? Does that make any sense to you?
Houses were wildly, astonishingly, incredibly pricey for 2 years -- to everyone EXCEPT the shills at the NAR!
Posted by: Jay Walker | Aug 13, 2008 2:11:43 PM
This is great to go after the way the NAR influences thinking on these questions. They are a lobby of sales folk, and do tend to preserve that aspect of what they do.
Is the concept of median, or average, in relation to a business like house sales very useful at all really? Wouldn't some kind of distribution function be much more useful for analysis?.
Most likely the median priced house doesn't exist something that doesn't exist, and the family which could buy it doesn't exist either. Useful for hypothetical thinking perhaps, such constructs may not be much use in looking at economic situations where the devil always seems to be in the details.
Housing is a segmented market. The segments are defined relative to income. Builders build different kinds of homes for people in different income brackets. The brackets continue over into the used home segment and provide the basis for price ranges. Condo, town, home, single family, the median price represents all of these different types of housing and different segments of the market. Wouldn't it be more useful to know what kinds of housing are selling rather than just tracking interest rates? Wouldn't a statistical distribution function which correlates type of housing and price with a similar type of distribution analysis on the income and asset side be more useful? I doubt the NAR could produce such data. Hope you don't mind this. Best wishes for what you are doing with this kind of work.
Posted by: Chris | Aug 13, 2008 2:27:48 PM
Thanks, Barry. My wife and I finally came around to the realization that it's better to rent in a neighborhood you can't afford to buy in than to "own" in one you can. It's all about proximity. Maybe someday--through a combination of income growth and price declines--I'll be able to buy in around here.
Posted by: Chris D. | Aug 13, 2008 2:48:50 PM
@Jay Walker: Not going to "flame" you as you suggest, but I do question just how many people can afford a 20% down payment at the current price levels. My wife and I have made solid salaries over the past 5 years+ (around $150-$225/year), don't spend extravagantly on "discretionary" items (don't carry any credit card balances), and it took us years to come up with a 20% down payment on a $376K home in the Twin Cities. I'm guessing those who make far less than that are having an even harder time coming up with 20% as well, especially in high cost areas on the coasts. Throw in the fact that many of these people do carry a lot of credit balances for discretionary purchases, and it becomes nearly impossible for them to come up with the 20%. Bottom line is too many people didn't have a down payment to offer and weren't required to put anything up then, but that's all changed now. That's simply a fact.
Posted by: Jeff | Aug 13, 2008 2:51:47 PM
I listened to your interview with Aaron Task today. While I can agree with your thesis, I did not care for the Love Canal-Detroit comment. Houses in the city of Detroit rarely go to $0 and certainly have some value to someone. This unfunny remark can probably be best summed up as racist or ignorant. I will assume the former.
~~~
BR: Nick, unfortunately, its not a joke:
http://bigpicture.typepad.com/comments/2008/02/detroit-housing.html
I will update this later today . . .
More updates: Foreclosure fallout: Houses go for a $1 http://www.detnews.com/apps/pbcs.dll/article?AID=/20080813/METRO/808130360/&imw=Y
Posted by: Nick | Aug 13, 2008 3:44:43 PM
Although this may seem like a detail (and it probably is), if you don't put 20% down you have to buy PMI (Privte mortgage insurance) which adds cost to the mortgage payment. Funny how that detail got ignored. Great article Barry.
Posted by: Carl Erickson | Aug 13, 2008 4:34:36 PM
Re Chris D and affordability of deposits. An income of 150-225K is 3-4 x the median. At the same level of expenditure as the median after tax a 75K deposit takes roughly a year to save. Even double the expenditure and it takes two.
Saving a deposit for a first home takes some denial and effort. Always has. Commonly it also means committing more than 25% of income to repayments for a few years. Even now with the average house price at 4-5 years income and interest at say 6.5% that is cheap housing compared with other places, less than half the costs in NZ for instance.
If you factor in the effect of rising land prices in the centres of cities and the increase in home sizes over the years ie changes in standards then part of the apparent decrease in affordability is from comparing apples and pears.
While actual affordability is influenced by other commitments eg debts, that largely reflects other priorities and a choice which is irrelevant as to whether housing could be afforded.
Similarly comparing rent with purchase is only really meaningful if it compares long term outcomes, namely increasing rents versus long term capital appreciation. I doubt over 10 or 20 years anyone could say they were better off renting, particularly as many would not in fact save and invest any lower outgoings they might have had.
Posted by: chris_gee | Aug 13, 2008 4:36:41 PM
